In this research, I investigated the Asymmetric Volatility Spillover Effects within and across six major international stock markets indexes namely United States (S&P 500), Canada (TSX 60), France (CAC 40), Germany (DAX 40), Italy (MIB 30) & Japan (NIKKEI 225) on 23 years of time series dataset (1999 - 2022)
My findings reveal that volatility spillovers are asymmetric, with the largest impact being auto-influence. Canada has the least impact on the other five countries, while Japan and the United States experience the least cross-influence volatility from abroad (most resilient). Italy is most impacted by external volatility changes. In order to realistically and accurately predict volatility, investors' expectations are a significant factor. And contrary to Bensaida (2019), the United States maintains dominance with respect to global market influence.
To investigate the Asymmetric Volatility Spillover Effects within and across International stock markets. Study design/methodology/approach GARCH (Generalized Autoregressive Conditional Heteroscedasticity) Model, Markov-switching model, and Vector Autoregressive Model.
Volatility spillovers are asymmetric, with the largest impact being auto-influence. Canada has the least impact on other countries, while Japan and the US experience the least cross-influence volatility from abroad. Italy is most impacted by external volatility changes. In order to realistically and accurately predict volatility, investors' expectations are a significant factor. In contrast to Bensaida (2019), the United States maintains dominance with respect to global market influence.
Existing Literature continues to use either realized volatility or conditional volatility without controlling for expected return of the market. This leads to overestimated volatility as well as biased spillover effects. This research combines GARCH and Markov-switching models to capture conditional volatility and expected return simultaneously. It also contributes to the good and bad volatility literature by showing that using time-varying expected return as threshold to distinguish good and bad is better than using 0.
Stakeholders in the financial sector (Hedge Funds, Money and Asset managers etc) can leverage findings and insights from this research to inform investment strategies, risk management and the exploitation of spillover opportunities. A reliable and robust data ecosystem, highly qualified analysts and organizational management are required for the implementation, productionization and deployment that leverages findings from this model.
Expected-Return, Asymmetry, Volatility, Spillover, GARCH, Markov-Switching, Vector Autoregression